Age and Ageing 1999; 28: 373–377 q 1999, British Geriatrics Society

Declining physical abilities with age: a

cross-sectional study of older twins and
centenarians in Denmark
KAREN ANDERSEN-RANBERG1, KAARE CHRISTENSEN1, BERNARD JEUNE1, AXEL SKYTTHE1, LONE VASEGAARD1,
JAMES W. VAUPEL1,2
1
Ageing Research Centre and Epidemiology, Institute of Public Health and 2The Danish Centre for Demographic
Research, Odense University, Winslowparken 17, DK-5000 Odense C, Denmark

Address correspondence to: K. Andersen-Ranberg. Fax: (þ45) 6590 6938. Email: K-Ranberg@win-chs.sdu.dk

Abstract
Objective: to evaluate whether physical disability reaches a plateau in the oldest age groups.
Design: cross-sectional survey.
Participants: a total of 3351 individuals, which included all those living in Denmark who celebrated their 100th
anniversary during the period from 1 April 1995 to 31 May 1996 (276 subjects) and all Danish twins aged 75–94
registered in the Danish Twin Register (3075 subjects).
Main outcome measures: the ability to perform selected items of basic activities of daily living independently.
Results: the prevalence of independence in each of six selected activities of daily living was significantly lower in
both men and women centenarians compared with octo- and septuagenarians. The sex difference in independence
in all six selected activities of daily living was larger for each advancing age group, with women being most disabled
(P < 0:001). In centenarians 20% of women and 44% of men were able to perform all selected activities of daily
living independently.
Conclusion: compared with individuals aged 75–79 years, physical abilities of men and women gradually
diminished in age groups 80–84, 85–90 and 90–94, with the lowest levels among 100-year-olds. Although women
have lower mortality, they are more disabled than men, and this difference is more marked with advancing age.

Keywords: activities of daily living, aged, centenarians, oldest-old, twins

Introduction challenge—if increased longevity is accompanied by

Mortality is levelling off in very old people [1]. In 12
countries with reliable demographic data, the number
of octogenarians has grown fourfold, nonagenarians
eightfold and centenarians more than 20-fold from
1950 to 1990 [2]. The dramatic increase in the number
of centenarians is largely the result of a decrease in
mortality among 80- and 90-year-olds [3]. At all ages,
mortality rates are lower in women than men, resulting
in a sex ratio (M : F) of about 1 : 3 at the oldest ages [2].
Physical ability is important for health and well
being in old age. It is influenced by the ageing process
and multiple diseases [4, 5]. With advancing age,
disability increases with a marked sex difference:
women over 75 generally have higher rates of disability
than men [6]. As disability can result in the need for
additional care, the increasing numbers of the oldest-
old could become a huge social and economic
an equal increase of life with disability [7]. However, it
is argued that increased longevity is accompanied by an
equal increase in years free from disability [8].
We have investigated whether physical ability
declines with age or reaches a plateau in the oldest-
old. We have also studied the sex differences in A
age-associated disability.
Materials and methods
The study is based on two nation-wide studies launched
by the Ageing Research Centre at Odense University:
the Longitudinal Study of Ageing in Danish Twins
(LSADT) and the Longitudinal Danish Centenarian
Study (DLCS).
LSADT, based on the Danish Twin Registry [9, 10],
was restricted to all registered Danish twins aged 75þ

373
K. Andersen-Ranberg et al.

before 1 January 1995 who had not emigrated instrument which has previously been used and
permanently [11]. In this study, 2383 of 3075 subjects validated in Denmark [12–16]. All the items refer to
aged 75–94 were interviewed at home by 100 what the participant was able to do on the day of the
interviewers from February to April 1995 (a participa- interview.
tion rate of 77.5%). A proxy respondent was sought We have chosen to focus on five most basic self-care
when the twin was unable to answer because of mental ADLs: (i) transferring from bed and chair, (ii) walking
incapacity or severe deafness. In this study we indoors, (iii) dressing the upper half of the body,
excluded those aged 95 years and older as the sample (iv) dressing the lower half and (v) using the toilet. To
size was small (24 people). compare these five basic self-care activities with a
The DLCS is a clinical epidemiological survey of all more demanding activity, we have included ‘walking
persons living in Denmark who celebrated their 100th outdoors in fine weather’.
birthday during the period 1 April 1995–31 May 1996
(276 people). The addresses of all centenarians in the Statistical methods
study population were obtained from the central
person registry using the personal identification Differences in type of housing between participants
number of each centenarian. About two weeks after and non-participants were tested with a x2-test. For
their 100th birthday, all centenarians received a letter each of six ADL items, a prevalence proportion ratio
explaining the study and asking permission for a (PPR) of remaining independent was calculated for
geriatrician and a geriatric nurse to visit them for an each age group using the 75–79-year-olds as a
interview and physical examination. A proxy respon- reference group. Confidence intervals around PPRs
dent was sought when the centenarian was unable to were calculated using the standard error of loge(PPR),
answer. A total of 207 (75%) centenarians (162 women and assuming normal distribution of loge(PPR). The
and 45 men) participated. centenarians were also compared separately with octo-
and nonagenarians. The gender difference between age
groups in the ability of performing all the six selected
Non-participants ADL items independently was tested with the Mantel–
Haenszel test.
In both studies all non-participants, including those
who died before the interview, were registered by age,
sex and type of housing.
Results
There were only minor differences in the response
Assessment of activities of daily living (ADLs) rates in the different age groups. More men than
Both the LSADT and the DLCS assess ADL with an women participated (women: 75.6%, men: 80.8%,

Table 1. Ability of women to perform each of six selected activities of daily living (ADLs) independently
Age group (years)
...........................................................................................................................................................

75–79 80–84 85–89 90–94 100

ADL (n ¼ 615) (n ¼ 493) (n ¼ 326) (n ¼ 99) (n ¼ 162)
........................................................................................................................................................................................................................

Transferring from bed/chair

PPR (and 95% CI) 1 0.96 (0.93, 0.99) 0.89 (0.85, 0.94) 0.78 (0.70, 0.88) 0.68 (0.61, 0.76)
n 595 457 282 75 107
Walking indoors
PPR (and 95% CI) 1 0.97 (0.94, 0.99) 0.90 (0.86, 0.94) 0.78 (0.70, 0.87) 0.63 (0.56, 0.71)
n 598 463 284 75 99
Using the toilet A
PPR (and 95% CI) 1 0.96 (0.93, 0.99) 0.87 (0.83, 0.91) 0.70 (0.61, 0.81) 0.58 (0.50, 0.66)
n 591 453 272 67 90
Dressing upper half
PPR (and 95% CI) 1 0.93 (0.90, 0.97) 0.87 (0.83, 0.92) 0.72 (0.63, 0.82) 0.44 (0.36, 0.52)
n 590 442 272 68 68
Dressing lower half
PPR (and 95% CI) 1 0.93 (0.90, 0.96) 0.86 (0.81, 0.90) 0.69 (0.60, 0.80) 0.33 (0.26, 0.41)
n 586 437 266 65 51
Walking outdoors in fine weather
PPR (and 95% CI) 1 0.82 (0.77, 0.88) 0.60 (0.54, 0.68) 0.50 (0.39, 0.63) 0.33 (0.26, 0.43)
n 524 344 168 42 46

PPR, prevalence proportion ratio (youngest age group as reference); 95% CI, 95% confidence interval.

374
 Older twins and centenarians in Denmark

Table 2. Ability of men to perform each of six selected activities of daily living (ADLs) independently
Age group (years)
...........................................................................................................................................................

75–79 80–84 85–89 90–94 100

ADL (n ¼ 358) (n ¼ 291) (n ¼ 168) (n ¼ 30) (n ¼ 45)
........................................................................................................................................................................................................................

Transferring from bed/chair

PPR (and 95% CI) 1 0.96 (0.92, 1.00) 0.97 (0.93, 1.02) 0.91 (0.79, 1.05) 0.84 (0.72, 0.97)
n 342 267 156 26 36
Walking indoors
PPR (and 95% CI) 1 0.96 (0.92, 1.00) 0.98 (0.94, 1.03) 0.94 (0.83, 1.06) 0.83 (0.72, 0.97)
n 343 268 158 27 36
Using the toilet
PPR (and 95% CI) 1 0.97 (0.93, 1.02) 0.95 (0.90, 1.01) 0.89 (0.75, 1.04) 0.78 (0.65, 0.93)
n 336 266 150 25 33
Dressing upper half
PPR (and 95% CI) 1 0.96 (0.92, 1.01) 0.95 (0.89, 1.01) 0.90 (0.77, 1.06) 0.62 (0.49, 0.80)
n 331 259 147 25 26
Dressing lower half
PPR (and 95% CI) 1 0.94 (0.89, 1.00) 0.93 (0.86, 0.99) 0.83 (0.68, 1.02) 0.60 (0.46, 0.79)
n 329 252 143 23 25
Walking outdoors in fine weather
PPR (and 95% CI) 1 0.88 (0.82, 0.96) 0.81 (0.73, 0.90) 0.63 (0.45, 0.88) 0.63 (0.47, 0.83)
n 305 219 116 16 24

PPR, prevalence proportion ratio (youngest age group as reference); 95% CI, 95% confidence interval.

P < 0:05). There were only small and statistically lower (P < 0:05) among centenarians compared to
insignificant differences in the proportion of septua-, octo- and nonagenarians, except for centenar-
people in nursing homes among participants and ian men compared to nonagenarians: here the same
non-participants. pattern was observed, but the sample size was small.
The proportions and PPRs of independence in For each independently-performed ADL, women con-
each of six selected ADLs are given for both women stituted a lower proportion than did men, and this
(Table 1) and men (Table 2) within each age group. proportional difference became more pronounced in
With advancing age a progressively lower proportion advancing age groups.
of women and men do each selected ADL item The sex difference was even more marked in the
independently. The proportion was significantly ability to carry out all six selected ADLs independently,
as shown in Figure 1. At ages 75–79 there was no
difference in abilities between men and women, while
there was a progressive and significant difference
between sex in advancing age groups (Mantel–
Haenszel test: P < 0:001). At age 75–79, 84% of all
women and 83% of all men were able to perform all six
selected basic ADLs, while the corresponding numbers
in centenarians were 20% and 44%.

Discussion A

In this cross-sectional study we found no evidence of a

levelling-off in physical disability at the most extreme
ages. On the contrary, we found a higher level of
disability with increasing age, including centenarians.
Furthermore, we found that the sex difference was
most pronounced in centenarians.
Non-medical trained interviewers carried out the
LSADT, while the DLCS used medically trained inter-
Figure 1. The proportions of men (B) and women (D), who viewers. However, this is unlikely to affect the result,
are able to carry out all selected activities of daily living because the measurement of basic ADL is robust [17].
independently, within each age group. The study benefited from a large sample size which

375
K. Andersen-Ranberg et al.
included individuals at the most extreme ages.
Furthermore, information on institutionalization sug-
gests that the participants were representative of
disability in the total population.
It has been suggested that twins have an increased
incidence of certain conditions, such as cardiovascular
and Alzheimer’s disease, which could pose a potential
threat to the validity of epidemiological research based
on twins [18, 19]. However, most twin–singleton
studies show that disease frequencies and mortality are
similar in twins and singletons [20–22]. The Danish
twin population is therefore likely to be representative
of the general Danish population of elderly people in
terms of health status and mortality.
The optimal way to evaluate changes in ADL
capabilities with advancing age is by longitudinal
study. But as only one in a hundred 80-year olds
reaches the age of 100 [23, 24], a study of the oldest-old
must include about 20 000 80-year-olds followed for 20
years to have a centenarian sample size comparable to
ours.
Because our data are cross-sectional, we are not
able to distinguish cohort from age effects. A cohort
effect could explain part of the age differences in
physical ability. However, in the Danish Glostrup
Population Study [25], no significant difference in
ADL capability was found between two 70-year-old
cohorts, born 1894 and 1914, while there was a
significant difference in cardiovascular risk factors, the
younger cohort being healthier. In contrast, the
National Long Term Care Survey, 1982–1994 [17]
found a reduction in the age-standardized rates of
disability among 5-year categories of 65–95þ year olds
over 12 years [26].
Period effects may influence the ADL pattern
observed, since there is evidence [2] that the decline
in oldest-old mortality is primarily caused by period
factors with immediate effects. If period effects, and
maybe even cohort effects, are present perhaps
centenarians of the future will be less disabled than
today, when 51% are in nursing homes and 67% are not
able to perform all basic ADLs.
Other studies have shown a marked sex difference
in functional capability in favour of men. In the EPESE
studies, there was a steeply increasing prevalence of
ADL disability with increasing age, and women
generally showed higher rates of disability than men
beginning around age 75 [27]. However, this study was
restricted to community dwellers, and the oldest were
pooled together in one group of 85þ year olds. The
Longitudinal Study of Physical Ability in the Oldest-Old
[5], derived from The United States National Health
Interview Survey, is also based on non-institutionalized
subjects aged 80 or older. In this study Harris et al.
found that a high level of physical ability was associated
with good health and a low risk of morbidity and
mortality, and men were more likely to remain at a high
level of physical ability in a 2-year follow-up. They also
376
found that women consistently reported more diffi-
culty with each physical activity.
Our data show that the sex difference in the ability
of performing basic ADL independently becomes
progressively larger as we move upwards in the age
groups and is in sharp contrast with the better survival
of women. This paradox has not yet been fully
explained. A differential mortality leading to a surviv-
ing group of highly selected men could partly explain
the better physical ability of the very old men. Another
explanation of the increasing sex difference in the
ability of performing basic ADLs could be the
differences in muscle mass between men and
women. Given that women are weaker and have
smaller muscles than men [28], and that people with
advancing age become less mobile and more inactive
leading to accelerated muscle atrophy [29, 30], a
threshold level for a minimum muscle mass require-
ment is reached earlier in women. This in turn further
increases disability and a vicious cycle perpetuates
[31], starting earlier in women. Perhaps if measures are
taken to maintain basic physical abilities, the need for
care could be reduced: even in very elderly people
exercise training is a feasible and effective means of
counteracting muscle weakness and physical frailty
[32].
Key points
• Cross-sectional data on 3351 Danes aged 75 or older
(including 276 centenarians) reveal that physical
ability was significantly lower in centenarians than
in octo- and septuagenarians.
• In the 75–79-year age group, 84% of all women and
83% of all men were able to perform all six selected
basic activities of daily living independently, while
the corresponding numbers among centenarians
were 20% and 44%.
• Intervention against physical inactivity in elderly
people may improve physical ability in extreme old
age.
Acknowledgement
This work was supported by The National Institute on A
Aging research grant NIA-PO1-AG08761 and The
Danish Interdisciplinary Research Council.
References
1. Vaupel JW, Carey JR, Christensen K et al. Biodemographic
trajectories of longevity. Science 1998; 280: 855–60.
2. Kannisto V. Development of Oldest-Old Mortality, 1950–1990:
evidence from 28 developed countries. Odense Monographs on
Population Aging 1. Odense: Odense University Press, 1994.
3. Vaupel JW, Jeune B. The emergence and proliferation of
 Older twins and centenarians in Denmark

centenarians. In: Jeune B, Vaupel JW eds. Exceptional Longevity: from whether diseases are genetically determined? Lancet 1993; 341:
the prehistory to the present. Odense Monographs on Population 1008–9.
Aging 2. Odense: Odense University Press, 1995.
20. Duffy DL. Twin studies in medical research. Lancet 1993; 341:
4. Guralnik JM, La Croix A, Abbott RD et al. Maintaining mobility in 1418–9.
late life. Demographic characteristics and chronic conditions. Am J
21. Christensen K, Vaupel J, Holm NV, Yashin AI. Mortality among
Epidemiol 1993; 137: 845–57.
twins after age 6: fetal origins hypothesis versus twin method. Br Med
5. Harris T, Kovar MG, Suzman R, Kleinman JC, Feldman JJ. J 1995; 310: 432–6.
Longitudinal study of physical ability in the oldest-old. Am J Public
22. Christensen K, Holm NV, Vaupel J. Alzheimer’s disease in twins.
Health 1989; 79: 698–702.
Lancet 1996; 347: 976.
6. Guralnik JM, Simonsick EM. Physical disability in older Americans.
23. Kannisto V. The Advancing Frontier of Survival: life tables for old
J Gerontol 1993; 48 (special issue): 3–10.
age. Odense Monographs on Population Aging 3. Odense University
7. Kramer M. The rising pandemic of mental disorders and associated Press, Odense, Denmark 1996.
chronic diseases and disabilities. Acta Psychiatr Scand 1980; 62
24. Jeune B, Kannisto V. Emergence of centenarians and super-
(suppl. 285): 382–97.
centenarians. In: Robine J-M, Vaupel JW, Jeune B, Allard M eds.
8. Manton KG. Changing concepts of morbidity and mortality in the Longevity: to the limits and beyond. Research and Perspectives in
elderly population. Milbank Quart 1982; 60: 183–244. Longevity. Foundation IPSEN. Berlin: Springer-Verlag, 1997.
9. Hauge M, Harvald B, Fischer M et al. The Danish Twin Register. 25. Jürgensen KS, Avlund K, Schroll M. The social situation in two 70-
Acta Genet Med Gemellol 1968; 2: 315–31. year-old populations. A cohort comparison of persons born in 1897
and 1914 cohorts from the Glostrup Population Studies (in Danish).
10. Kyvik KO, Christensen K, Skytthe A, Harvald B, Holm NV. The
Ugeskr Laeger 1990; 152: 2993–7.
Danish Twin Registry. Danish Med Bull 1996; 43: 467–70.
26. Manton KG, Stallard E, Corder LS. The dynamics of dimensions of
11. McGue M, Christensen K. Genetic and environmental contribu-
age-related disability 1982 to 1994 in the US elderly population.
tions to depression symptomatology: evidence from Danish twins 75
J Gerontol A Biol Sci Med Sci 1998; 53(1): B59–70.
years and older. J Abnorm Psychol 1997; 106: 439–48.
27. Cornoni-Huntley J, Brock DB, Ostfeld A, Taylor JO, Wallace RB,
12. Avlund K, Schultz-Larsen K. What do 70 year-old men and
eds. Established Populations for Epidemiologic Studies of the Elderly,
women do? And what are they able to do? From the Glostrup survey
Resource Data Book. NIH publication no. 86–2443. Bethesda, MD:
in 1984. Aging Clin Exp Res 1991; 3: 39–49.
National Institutes of Health, 1986.
13. Schultz-Larsen K, Avlund K, Kreiner S. Functional ability of
28. Lexell J. Human aging, muscle mass, and fiber type composition.
community dwelling elderly. Criterion-related validity of a new
J Gerontol 1995; 50A (special issue): 11–6.
measure of functional ability. J Clin Epidemiol 1992; 45: 1315–26.
29. Rodgers MA, Evans WJ. Changes in skeletal muscle with aging:
14. Avlund K, Kreiner S, Schultz-Larsen K. Construct validation and
effects of exercise training. Exerc Sports Sci Rev 1993; 21: 65–102.
the Rasch model: functional ability of elderly people. Scand J Soc Med
1993; 21: 233–45. 30. Coggan AR, Spina RJ, King DS et al. Histochemical and enzymatic
comparison of the gastrocnemius muscle of young and elderly men
15. Avlund K, Kreiner S, Schultz-Larsen K. Functional ability scale for
and women. J Gerontol Biol Sci 1992; 47: B71–6.
the elderly. Eur J Publ Health 1996; 6: 35–42.
31. Tseng BS, Marsh DR, Hamilton MT, Booth FW. Strength and
16. Olsen H, Jeune B, Andersen-Ranberg K. Centenarians in a county
aerobic training attenuate muscle wasting and improve resistance to
of Denmark. Morbidity and functional capacity (in Danish). Ugeskr
the development of disability with aging. J Gerontol 1995; 50A
Laeger 1996; 158: 7397–401.
(special issue): 113–9.
17. Manton KG, Corder L, Stallard E. Chronic disability trends in
32. Fiatarone MA, O’Neill EF, Ryan ND et al. Exercise training and
elderly United States populations: 1982–1994. Proc Natl Acad Sci
nutritional supplementation for physical frailty in very elderly
USA 1997; 94: 2593–8.
people. N Engl J Med 1994; 330: 1769–75.
18. Räihä I, Kaprio J, Koskenvuo M, Rajala T, Sourander L.
Alzheimer’s disease in Finnish twins. Lancet 1996; 347: 573–8.
Received 8 December 1997; accepted in revised form 22 July
19. Phillips DIW. Twins studies in medical research: can they tell us 1998

377